A Machine Learning-Driven Smart Livestock Monitoring and Risk-Stratified Health Management System: A Prototype Evaluated on Simulated Data

Late detection of disease is a persistent problem for cattle keepers in Nigeria, where veterinary visits are few and most health checks still rely on what a farmer notices. This paper describes a prototype health-monitoring system that places each animal in one of four risk tiers (low, medium, high or critical) using 15 indicators that a farm worker can record without laboratory tests and compares six classifiers on the same task.